• Login
    View Item 
    •   USU-IR Home
    • Faculty of Computer Science and Information Technology
    • Department of Information Technology
    • Master Theses
    • View Item
    •   USU-IR Home
    • Faculty of Computer Science and Information Technology
    • Department of Information Technology
    • Master Theses
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    “Modifikasi Algoritma Apriori dengan Substitusi Metode Bayesian pada Nilai Confidence Terhadap Aturan Asosiasi

    View/Open
    fulltext (2.594Mb)
    Date
    2018
    Author
    Ginting, Dewi Sartika
    Advisor(s)
    Mawengkang, Herman
    Efendi, Syahril
    Metadata
    Show full item record
    Abstract
    Apriori Algorithm one of the data mining algorithms in the formation of association rule mining. A priori algorithm is the process of extraction of information from a database, followed by frequent item / itemset and candidate generation in formation of association rule mining in order to obtain minimum value of support and minimum confidence value. The value of confidence has a big effect on the resulting rule, where the rule generated by the k-itemsets pattern needs to be calculated on the level of confidence or certainty of the k-itemsets pattern that has complied with the rules. Therefore, this research discusses about a priori algorithm modification which focuses on giving confidence value for each rule generated. Modifications are made by substituting the Bayesian method on a standard Apriori confidence formula. But at the beginning of the pre-processing data is done item reduction, due to the number of items that are very small value of occurrence in sales transactions. In the process of item reduction, the author uses Principal Component Analysis method. From the result of research done for pre-processing, there is item reduction equal to 77,7% and for next process there is difference of confidence value between standard apriori and modification, where the value of confidence generated a priori modification is bigger, and after calculated for some rules taken according to the minimum requirement of support, there is an average difference of confidence value of 6.81%.
     
    Algoritma Apriori salah satu algoritma data mining dalam pembentukan asosiasi rule mining. Algoritma apriori adalah proses ekstraksi informasi dari suatu database, dilanjutkan dengan melakukan frequent item/itemset dan candidate generation dalam pembentukan asosiasi rule mining guna mendapatkan hasil nilai minimum support dan nilai minimum confidence. Nilai confidence berpengaruh besar terhadap rule-rule yang dihasilkan, dimana rule yang dihasilkan oleh pola kitemsets perlu di hitung nilai tingkat kepercayaan atau kepastian dari pola kitemsets yang sudah memenuhi aturan. Untuk itu penelitian ini membahas tentang modifikasi algoritma apriori yang terfokus pada pemberian nilai confidence untuk setiap rule yang dihasilkan. Modifikasi yang dilakukan dengan mensubtitusi metode bayesian pada formula confidence di apriori standart. Namun di awal pada pre-processing data dilakukan reduksi item, dikarenakan banyaknya item-item yang sangat kecil nilai kemunculan dalam transaksi penjualan. Pada proses reduksi item, penulis menggunakan metode Principal Componen Analysis. Dari hasil penelitian yang dilakukan untuk pre-processing maka terjadi reduksi item sebesar 77,7% dan untuk proses selanjutnya terdapat perbedaan nilai confidence antara apriori standar dan modifikasi, dimana nilai confidence yang dihasilkan apriori modifikasi lebih besar, dan setelah dihitung untuk beberapa aturan yang diambil sesuai ketentuan minimum support maka terdapat rata-rata perbedaan nilai confidence sebesar 6,81%.

    URI
    http://repositori.usu.ac.id/handle/123456789/4942
    Collections
    • Master Theses [623]

    Repositori Institusi Universitas Sumatera Utara - 2025

    Universitas Sumatera Utara

    Perpustakaan

    Resource Guide

    Katalog Perpustakaan

    Journal Elektronik Berlangganan

    Buku Elektronik Berlangganan

    DSpace software copyright © 2002-2016  DuraSpace
    Contact Us | Send Feedback
    Theme by 
    Atmire NV
     

     

    Browse

    All of USU-IRCommunities & CollectionsBy Issue DateTitlesAuthorsAdvisorsKeywordsTypesBy Submit DateThis CollectionBy Issue DateTitlesAuthorsAdvisorsKeywordsTypesBy Submit Date

    My Account

    LoginRegister

    Repositori Institusi Universitas Sumatera Utara - 2025

    Universitas Sumatera Utara

    Perpustakaan

    Resource Guide

    Katalog Perpustakaan

    Journal Elektronik Berlangganan

    Buku Elektronik Berlangganan

    DSpace software copyright © 2002-2016  DuraSpace
    Contact Us | Send Feedback
    Theme by 
    Atmire NV